It is well known that the capacity of a capacitor is related to the quality of dynamic random access memory (DRAM). Therefore, many methods have been developed for increasing the capacity of capacitors.
First of all, please refer to FIGS. 1(a).about.(b) showing a conventional method for manufacturing a capacitor. This method is described as follows.
Shown in FIG. 1(a) are the steps of (1) forming an interlayer dielectric (ILD) 11 over a silicon substrate 10 by chemical vapor deposition (CVD) or lower pressure chemical vapor deposition (LPCVD), (2) defining a contact window by photolithography and partially removing the ILD 11 to form the contact window 12, (3) forming a doped polysilicon layer 13 with a thickness of 1000 .ANG. over the ILD 11 and in the contact window 12 by LPCVD, and (4) forming a rugged polysilicon layer 14 with a thickness of 850 .ANG. over the doped polysilicon layer 13 to increase the surface area of the capacitor.
In FIG. 1(b), the steps include: 1) defining the capacitor region by photolithography and partially etching the rugged polysilicon layer 14 and the doped polysilicon layer 13 to expose a portion of the ILD 11; 2) forming an oxide-on-nitride-on-oxide (ONO) layer 15 on the rugged polysilicon layer 14 and the ILD 11 and alongside the doped polysilicon layer 13 by LPCVD; and 3) forming another doped polysilicon layer 16 over the ONO layer 15 to construct the conventional capacitor.
Please refer to FIG. 2 showing another conventional method. The detailed steps are illustrated as follows.
In FIG. 2(a), the steps include: (1) forming an interlayer dielectric (ILD) 21 over a silicon substrate 20 by chemical vapor deposition (CVD), (2) forming a silicon nitride 22 on ILD 21, wherein the silicon nitride 22 has a thickness of 100 .ANG..about.300.ANG. and serves as an etching stop layer, (3) forming a sacrificial oxide 23 on the silicon nitride 22 by CVD, (4) defining a contact window by photolithography and partially removing the ILD 21, the silicon nitride 22, and the sacrificial oxide 23 to form the contact window 24, and (5) forming a doped polysilicon layer 25 with a thickness of 1000 .ANG. over the sacrificial oxide 23 and in the contact window 24 by LPCVD.
In FIG. 2(b), the steps include: 1) defining the capacitor region by photolithography and partially etching the doped polysilicon layer 25; 2) etching the sacrificial oxide 23 by using a buffer over etching (B.O.E.) solution containing hydrofluoric acid (HF) to expose the silicon nitride 22; 3) forming an oxide-on-nitride-on-oxide (ONO) layer 26 on the silicon nitride 22 and a top and sidewalls of the doped polysilicon layer 25 by LPCVD; and 4) forming another doped polysilicon layer 27 on the ONO layer 26 to construct the capacitor.
In addition, there is another method as shown in FIG. 3. This method is described as follows.
In FIG. 3(a), the steps include: (1) forming an interlayer dielectric (ILD) 31 over a silicon substrate 30 by CVD, (2) forming a silicon nitride 32 on ILD 31, wherein the silicon nitride 32 has a thickness of 100 .ANG..about.300.ANG. and serves as an etching stop layer, (3) forming a first sacrificial oxide 33 on the silicon nitride 32 by CVD, (4) defining a contact window by photolithography and partially removing the first sacrificial oxide 33, the silicon nitride 32, and the ILD 31 to form the contact window 34, (5) forming a first doped polysilicon layer 35 with a thickness of 1000 .ANG. over the first sacrificial oxide 33 and in the contact window 34 by LPCVD, and (6) forming a second sacrificial oxide 36 on the first doped polysilicon layer 35 by CVD.
In FIG. 3(b), the steps include: 1) defining the capacitor region by photolithography and partially etching the second sacrificial oxide 36, the first doped polysilicon layer 35, and the first sacrificial oxide 33, wherein the silicon nitride 32 serves as an etching stop layer; 2) forming a second doped polysilicon layer 37 on the top surface of the second sacrificial oxide 36, alongside the second sacrificial oxide 36, the first doped polysilicon layer 35, and the first sacrificial oxide 33, as well as on the silicon nitride 32.
In FIG. 3(c), the second doped polysilicon layer 37 is etched by dry etching (i.e. an anisotropic etching) to expose the top surface of the second sacrificial oxide 36 and a portion of the silicon nitride 32.
In FIG. 3(d), the second sacrificial oxide 36 is completely removed by using a buffer over etching (B.O.E.) solution containing hydrofluoric acid (HF) to expose the first doped polysilicon layer 35. Thereafter, an oxide-on-nitride-on-oxide (ONO) layer 38 is formed over the portion of the silicon nitride 32, the second doped polysilicon layer 37, and the first doped polysilicon layer 35 by LPCVD. Finally, another doped polysilicon layer 39 is formed on the ONO layer 38 to construct the capacitor.
However, these conventional methods have some defects described as follows:
1. In FIGS. 1(a) and (b), the rugged polysilicon layer in the fixed capacitor region cannot effectively increase the surface area of the capacitor. Therefore, the maximum capacity can be only increased up to two times by such a method using the rugged polysilicon layer for increasing the surface area of capacitor. Because the size of the capacitor will be getting smaller in the future, this method may be no longer effective then.
2. In the method of FIGS. 2(a) and (b), the sacrificial oxide is formed and then is etched in order to increase the surface area of the capacitor, but the effect is not good enough.
3. In the method as shown in FIGS. 3(a).about.(d), the cylindrical doped polysilicon can increase the surface area of the capacitor which is constructed by a doped polysilicon layer, the ONO layer, and another doped polysilicon layer. However, it can be seen from FIG. 3(d) that the surface of the capacitor is so irregular that it will seriously influence the subsequent planarization process of the semiconductor.
Therefore, the present invention is developed to improve the above-described disadvantages.